Super-resolution refers to higher resolution images constructed from lower resolution originals. In video, different frames may be regarded as different representations of the same scene, and therefore, it is possible to have a better picture of the scene by fusing at least two adjacent frames.
Conventional methods obtain super resolution by processing video data in the pixel domain. For instance, when a decoder receives compressed video data, the decoder of the conventional art processes the compressed video data in two steps: 1) decompress the video data, and 2) process the decompressed video to obtain super-resolution. However, the two step conventional process in high complexity.